The tailoring of the structure and properties of singlemolecule magnets (SMMs) is a very active branch of modern coordination chemistry. Investigations on classes of structurally related SMMs, such as those of the Mn12 family, have helped unravel the mechanisms underlying slow magnetic relaxation in high-spin molecules, a key step in both fundamental and application-oriented research. At the same time, the search for new SMMs with higher blocking temperatures has fueled synthetic efforts aimed, on one side, at increasing structural control on cluster architectures and, on the other side, at developing the so-called serendipitousassembly approach. An elegant strategy for structural design is based on site-specific modification of preformed clusters. Carboxylate abstraction from Mn12 clusters, for instance, has been carried out site specifically using a variety of incoming ligands, such as nitrates, phosphanates, phosphates, or different carboxylates. However, these substitutions are accompanied by only small perturbation of the magnetic properties. The SMM behavior is associated with the magnetic anisotropy of clusters, which in turn depends on local anisotropies and on the way they vectorially add to give a resulting total anisotropy. Herein we show that site-specific ligand replacement provides a means to raise the symmetry of Fe4 clusters fromC2 toD3, which results in a dramatic increase of magnetic anisotropy and energy barrier. Fe4 clusters are among the simplest inorganic systems showing SMM behavior. 6] The archetypal member of this class is the tetrairon(iii) compound [Fe4(OMe)6(dpm)6] (1) (Hdpm= dipivaloylmethane). The six m-methoxide ligands bridge a central iron(iii) ion to three peripheral iron centers arranged at the vertices of an isosceles triangle with a crystallographic C2 symmetry, and some disorder which yields three different isomers in the lattice. At low temperature, the cluster has a high-spin state (S= 5) and an easy-axis magnetic anisotropy, two requisites for the observation of slowmagnetic relaxation. For the major component, the second-order zero-field splitting (ZFS) parameters are D= 0.206(1) cm 1 and E= 0.010(3) cm . In addition, the presence of sizeable fourth-order contributions has been demonstrated. To eliminate the lattice disorder, we attempted to replace the methoxide bridges in 1 with a tripodal ligand, 1,1,1tris(hydroxymethyl)ethane (H3thme), which affords facial coordination in octahedral iron(iii) complexes [Eq. (1)].